Seb Pereira specialises in machine learning, computer vision and multi-modal sensor fusion. His work focuses on evaluating the capabilities and limitations of modern AI models and sensing technologies, particularly in complex and constrained environments.
At OTUA, Seb conducts AI training, evaluation and testing, supporting the development of the voice AI and digital automation solution. He works across data preparation, prompt and model optimisation, and system validation to ensure solutions are reliable, accurate and aligned with real-world operational requirements. His role includes designing and refining conversational flows, implementing and testing integrations, analysing system performance, and iterating on models to improve patient engagement and operational efficiency. He works closely with both technical and operational stakeholders to translate requirements into practical solutions that are robust, compliant and scalable.
In parallel, Seb conducts research in object detection and sensor fusion, exploring how combinations of optical, infrared and depth sensing can be used to improve detection capability in nuclear glovebox environments. His work focuses on proof-of-concept systems and understanding performance limits across sensing modalities.